design-system

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The skill extracts style data from external target URLs provided by the user. This is conducted through the author's own headless agent tools (crawlio-agent-headless), which is the primary intended functionality of the skill.
  • [COMMAND_EXECUTION]: Utilizes the mcp__crawlio__call tool to invoke remote routines like skill_extract_design_system and skill_clone_site. These commands are scoped to the vendor-controlled environment.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites, creating a surface for indirect prompt injection. 1. Ingestion points: The skill_extract_design_system tool reads computed styles from the DOM of target URLs. 2. Boundary markers: Absent. 3. Capability inventory: Includes remote tool invocation and code fragment generation. 4. Sanitization: Absent. The risk is mitigated by the structured nature of the extracted data (CSS properties) rather than natural language or executable scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 01:06 AM
Security Audit — agent-trust-hub — design-system